Volume 98, Issue 2 pp. 273-285

Research Article

Open Access

Association Between Vascular NOTCH3 Aggregation and Disease Severity in a CADASIL Cohort – Implications for NOTCH3 Variant-Specific Disease Prediction

Minne N. Cerfontaine MD,

Minne N. Cerfontaine MD

orcid.org/0000-0001-9357-270X

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Gido Gravesteijn MD, PhD,

Gido Gravesteijn MD, PhD

orcid.org/0000-0002-2610-8728

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Remco J. Hack MD, PhD,

Remco J. Hack MD, PhD

orcid.org/0000-0001-9302-5802

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Kyra L. Dijkstra BSc,

Kyra L. Dijkstra BSc

Department of Pathology, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Mar Rodríguez-Girondo PhD,

Mar Rodríguez-Girondo PhD

Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Benno Gesierich PhD,

Benno Gesierich PhD

Medical Image Analysis Centre (MIAC) and Department of Biomedical Engineering, University of Basel, Basel, Switzerland

Search for more papers by this author

Marie-Noëlle W. Witjes-Ané PhD,

Marie-Noëlle W. Witjes-Ané PhD

Department of Geriatrics and Department of Psychiatry, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Remco van Doorn MD, PhD,

Remco van Doorn MD, PhD

Department of Dermatology, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Marco Duering MD,

Marco Duering MD

Medical Image Analysis Centre (MIAC) and Department of Biomedical Engineering, University of Basel, Basel, Switzerland

Search for more papers by this author

Julie W. Rutten MD, PhD,

Julie W. Rutten MD, PhD

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

These authors contributed equally to this work.

Search for more papers by this author

Saskia A. J. Lesnik Oberstein MD, PhD,

Corresponding Author

Saskia A. J. Lesnik Oberstein MD, PhD

[email protected]

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

These authors contributed equally to this work.

Address correspondence to Dr Saskia A. J. Lesnik Oberstein, Department of Clinical Genetics, Leiden University Medical Center, Albinusdreef 2, 2300 RC Leiden, The Netherlands. E-mail: [email protected]

Search for more papers by this author

Minne N. Cerfontaine MD,

Minne N. Cerfontaine MD

orcid.org/0000-0001-9357-270X

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Gido Gravesteijn MD, PhD,

Gido Gravesteijn MD, PhD

orcid.org/0000-0002-2610-8728

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Remco J. Hack MD, PhD,

Remco J. Hack MD, PhD

orcid.org/0000-0001-9302-5802

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Kyra L. Dijkstra BSc,

Kyra L. Dijkstra BSc

Department of Pathology, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Mar Rodríguez-Girondo PhD,

Mar Rodríguez-Girondo PhD

Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Benno Gesierich PhD,

Benno Gesierich PhD

Medical Image Analysis Centre (MIAC) and Department of Biomedical Engineering, University of Basel, Basel, Switzerland

Search for more papers by this author

Marie-Noëlle W. Witjes-Ané PhD,

Marie-Noëlle W. Witjes-Ané PhD

Department of Geriatrics and Department of Psychiatry, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Remco van Doorn MD, PhD,

Remco van Doorn MD, PhD

Department of Dermatology, Leiden University Medical Center, Leiden, The Netherlands

Search for more papers by this author

Marco Duering MD,

Marco Duering MD

Medical Image Analysis Centre (MIAC) and Department of Biomedical Engineering, University of Basel, Basel, Switzerland

Search for more papers by this author

Julie W. Rutten MD, PhD,

Julie W. Rutten MD, PhD

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

These authors contributed equally to this work.

Search for more papers by this author

Saskia A. J. Lesnik Oberstein MD, PhD,

Corresponding Author

Saskia A. J. Lesnik Oberstein MD, PhD

[email protected]

Department of Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands

These authors contributed equally to this work.

Search for more papers by this author

First published: 23 April 2025

https://doi.org/10.1002/ana.27240

Share a link

Email
Wechat
Bluesky

Abstract

Objective

Vascular NOTCH3 protein ectodomain aggregation is a pathological hallmark of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), a monogenic small vessel disease typically caused by cysteine-altering variants in NOTCH3. Given their high population frequency, these NOTCH3 variants are an important genetic contributor to stroke and vascular dementia worldwide. Disease severity in CADASIL is highly variable and is mainly determined by the position of the pathogenic NOTCH3 variant in the NOTCH3 ectodomain. Here, we aimed to investigate the association between NOTCH3 aggregation load in skin vessels, cysteine-altering NOTCH3 variants, and disease severity in a prospective cohort study of 212 patients with CADASIL with 39 distinct cysteine-altering NOTCH3 variants.

Methods

NOTCH3 aggregation load in skin vessels was determined by calculating the NOTCH3 score; the fraction of skin vessel wall area positive for NOTCH3 staining. Variant-specific NOTCH3 scores were calculated for variants present in 10 or more participants, by averaging the NOTCH3 scores of individuals with that distinct variant. The associations between the NOTCH3 score, NOTCH3 variants, and neuroimaging and clinical outcomes were investigated using multivariable linear mixed models, Cox regression, and mediation analyses.

Results

The NOTCH3 score was significantly associated with lifetime stroke probability and small vessel disease neuroimaging outcomes, but not with age. Variant-specific NOTCH3 scores reflected differences in disease severity associated with distinct NOTCH3 variants.

Interpretation

These findings suggest that differences in NOTCH3 aggregation propensity underlie the differences in disease severity associated with NOTCH3 cysteine-altering variants, and show that NOTCH3-variant specific NOTCH3 scores can contribute to improved individualized disease prediction in CADASIL. ANN NEUROL 2025;98:273–285

Aggregation of the ectodomain of the NOTCH3 protein (NOTCH3^ECD) in small- to medium-sized brain arteries is a major driver of vascular pathology in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL).¹ CADASIL is typically caused by cysteine-altering missense variants in NOTCH3, which lead to aberrant disulfide bridge formation in one of the 34 epidermal growth factor-like repeat (EGFr) domains of the NOTCH3^ECD.^2-5 This causes NOTCH3^ECD multimerization and aggregation in the media of the vessel wall, with sequestration of extracellular matrix proteins, such as HTRA1, TIMP3, and vitronectin.^{1, 6, 7} CADASIL disease severity is highly variable,^8-12 and cysteine-altering NOTCH3 variants (NOTCH3^cys) are an important contributor to both early- and late-onset stroke and dementia worldwide.^{8-10, 13, 14}

Vascular NOTCH3^ECD aggregation can be visualized as granular NOTCH3^ECD-positive staining using immunohistochemistry, or as granular osmiophilic material (GOM) using electron microscopy.^{15, 16} Although clinical signs and symptoms of CADASIL are confined to the brain, NOTCH3^ECD aggregation is present in small arteries in all organs, including the skin. NOTCH3^ECD aggregation load in skin vessels has been shown to correlate with NOTCH3^ECD aggregation load in brain vessels.¹⁶

Hallmark symptoms in patients with CADASIL include mid-adult onset of ischemic strokes and cognitive decline, neuropsychiatric disturbances such as apathy and depression, and migraine with aura. Brain magnetic resonance imaging (MRI) features are white matter hyperintensities (WMHs), lacunes, perivascular spaces (PVSs), cerebral microbleeds (CMBs), and brain atrophy.¹⁷ CADASIL disease severity is strongly associated with the position of the NOTCH3 variant along one of 34 EGFr domains of the NOTCH3 protein.^18-25 NOTCH3^cys are stratified into 3 risk categories for developing severe disease, based on odds ratios of patient-to-population frequencies of the EGFr domain in which they are located: low-risk (LR-NOTCH3), moderate-risk (MR-NOTCH3), or high-risk (HR-NOTCH3) variants.²⁰

Recently, in small sample sizes, HR-NOTCH3 variants have been shown to be associated with a higher NOTCH3^ECD aggregation load than MR- and LR-NOTCH3 variants,^{16, 20} suggesting a relation between vascular NOTCH3^ECD aggregation load and disease severity. A direct association between NOTCH3^ECD aggregation load and disease severity has never been studied. Understanding the interplay between NOTCH3 variants, NOTCH3^ECD aggregation, and disease severity is particularly relevant in the light of the current development of disease modifying therapies targeting vascular NOTCH3 aggregates.^{26, 27} In this prospective study of 212 individuals with a NOTCH3^cys variant from CADASIL pedigrees,^{18, 28} we investigated whether NOTCH3^ECD aggregation load in skin vasculature is associated with disease severity, and with distinct NOTCH3^cys variants.

Methods

Study Participants

Participants of the prospective CADASIL cohort Disease Variability in NOTCH3-associated Small Vessel Disease (DiViNAS) were included, consisting of patients with CADASIL and premanifest relatives with a NOTCH3^cys variant. A table of the 39 distinct NOTCH3 variants in DiViNAS participants can be found in Supplementary Table S1. Details concerning participant inclusion and the study protocol have been published elsewhere.^{18, 28} The study was approved by the medical ethics committee Leiden-The Hague-Delft (P21.013, P18.164, and P17.170). All participants gave written informed consent and procedures were carried out in accordance with the Declaration of Helsinki.

All data from study participants included in DiViNAS was collected during 1 or 2 study visits (at baseline and at a follow-up visit 2 years later). In addition to brain MRI, medical history, and neuropsychological testing, a 4-mm skin punch biopsy was taken from the posterior upper arm. In addition, brain tissue of 12 deceased patients with CADASIL was collected. This paper follows the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) guidelines.²⁹

Skin Biopsies and NOTCH3^ECD Immunohistochemistry

A total of 344 skin biopsies were taken of 218 study participants. Twenty-two skin biopsies did not fulfill the criterion of having at least 20 blood vessels per slide. Two hundred twelve participants (HR-NOTCH3 n = 117, MR-NOTCH3 n = 88, LR-NOTCH3 n = 7) with at least 1 skin biopsy were included in the study, of which 110 also had a second skin biopsy taken at the 2-year follow-up time point. Skin biopsies were formalin fixed and embedded in paraffin. Per tissue specimen, two 5-μm sections with a slice interval of 25 μm were pretreated with 0.1% trypsin for 30 minutes at 37°C and washed 3 times with phosphate-buffered saline (PBS). The slides were then incubated for 2 hours at room temperature with a primary mouse anti-NOTCH3^ECD antibody (clone 1E4, Millipore, dilution 1:1000, RRID:AB_2890101). A 2-step detection system (Brightvision, ImmunoLogic, VWRKC-DPVB55HRP) was used per the manufacturer's protocol. In short, the slides were post-blocked for 15 minutes, washed with PBS, incubated for 30 minutes with the ready-to-use goat anti-mouse/rabbit HRP antibody, washed in PBS, and subsequently stained with 3,3′-Diaminobenzidine (DAB) + Substrate Chromogen System (Dako, K3468, diluted 1:50). Counterstaining was then performed with 1:10 diluted Harris Hematoxylin for 10 seconds. The slides were stained in 5 batches and the samples were randomly distributed to minimize batch effects. For each patient, baseline and follow-up sections were stained in the same batch. In addition to skin, brain tissue of deceased patients with CADASIL with HR- (n = 9) and MR-NOTCH3 (n = 3) variants was sequentially sectioned with an interval of 5 μm and stained for NOTCH3^ECD according to the same protocol. Characteristics of the brain donors can be found in the Supplementary Table S2. The slides were scanned using the PANNORAMIC 250 slide scanner (3DHISTECH Kft, Budapest, Hungary) with 52 × magnification and the full focus setting in order to capture all NOTCH3^ECD granules within the 5-μm thick slide.

Quantification of NOTCH3^ECD Aggregation

Quantification of NOTCH3^ECD aggregation (the NOTCH3 score) was performed as previously described.¹⁶ Briefly, vessel wall regions of interest were manually determined by 2 blinded observers (authors M.C. and G.G.) using QuPath,³⁰ and the NOTCH3-positive area was determined using Color Threshold in ImageJ (https://imagej.net/ij/) parameterized to capture NOTCH3 granules (hue 40–210 [stop], saturation 0–255, and brightness 0–150; Fig 1A). Next, the NOTCH3-positive area was divided by the total vessel area, and the average of the 10 vessels with the highest score was calculated per sample. For each study participant, baseline and follow-up samples were scored by the same observer. The inter-rater reliability for the NOTCH3 score was determined in 60 random samples rated by 2 observers blinded to each other's scores, and to the genotype and clinical information of participants (single fixed rater intraclass correlation coefficient: 0.95, p < 2.2 × 10⁻¹⁶). Additionally, a variant-specific NOTCH3 score (the average NOTCH3 score of participants with a distinct variant) was calculated for NOTCH3 variants present in n ≥ 10 participants.

Details are in the caption following the image — **FIGURE 1**
Open in figure viewer PowerPoint

The association between the NOTCH3 score in skin vasculature with clinical and neuroimaging outcomes and with age. (A) Examples of NOTCH3 score measurements in two blood vessels. (1) Schematic overview of a skin biopsy, with the dermal and subcutaneous layers in gray and the epidermal layer in green. (2) Original images of blood vessels stained with a NOTCH3^ECD-antibody (*upper panel* scale bar = 10 μm and *lower panel* scale bar = 20 μm). (3) Region of interest drawn in pink (the blood vessel wall). (4) Thresholding of NOTCH3^ECD positive area. The NOTCH3 score was calculated by dividing the NOTCH3^ECD-positive area by the total vessel area, and then taking the average for the 10 most highest scoring vessels per sample. (B–G) Scatterplots with the standardized NOTCH3 score on the x-axis (after square root transformation) and standardized neuroimaging outcomes on the y-axis (conditional Pearson's residuals derived from linear mixed effects models corrected for age, sex, cardiovascular risk factors, and batch effects). There was a significant correlation between the NOTCH3 score and PSMD (p = 6.0 × 10⁻¹¹, p_adj = 4.2 × 10⁻¹⁰ [B]), nWMHv (p = 5.4 × 10⁻¹², p_adj = 4.3 × 10⁻¹¹ [C]), nLV (p = 1.5 × 10⁻⁷, p_adj = 8.8 × 10⁻⁷ [D]), CMB count (p = 0.0017, p_adj = 0.0057 [E]), and PVS score (p = 0.0014, p_adj = 0.0057 [F]) but not with BPF (p = 0.72, p_adj = 1 [G]) after correction for age, sex, and cardiovascular risk factors and batch effects. (H) Kaplan–Meier plot showing the differences in stroke probability between the tertiles of the NOTCH3 score (with the first tertile containing those with the lowest score, and the third tertile the highest). A higher NOTCH3 score was associated with a higher lifetime stroke probability (p = 3.1 × 10⁻⁵, p_adj = 1.5 × 10⁻⁴ [H]), after correction for sex and cardiovascular risk factors and batch effects. There was no correlation between the NOTCH3 score and TMT_B/A (p = 0.74, p_adj = 1 [I]). (J–K) There was no association between age (range = 26–81 years) and the NOTCH3 score (p = 0.75, p_adj = 1 [J]). The NOTCH3 score of brain vessels was not correlated with the age of death (p = 0.54 [K]). PSMD = peak width of the skeletonized mean diffusivity; nWMHv = normalized white matter hyperintensity volume; nLV = normalized lacune volume; CMB = cerebral microbleed; PVS = perivascular space; BPF = brain parenchymal fraction; NOTCH3^ECD = NOTCH3 ectodomain; TMT_B/A = Trail Making Test B Given A t scores; p_adj = multiple correction adjusted p value (using the Holm method, n = 8 tested hypotheses for neuroimaging and clinical outcomes; n = 7 tested hypotheses for modifiers of the NOTCH3 score). Levels of significance: **** = p < 0.0001, *** = p < 0.001, ** = p < 0.01, * = p < 0.05, ns = not significant.

Clinical and Neuroimaging Measures

Clinical outcomes were stroke, defined as neurological deficit lasting >24 hours in the absence of other probable causes or a diagnosis of ischemic stroke in the medical history, and executive function assessed by the Trail Making Test (TMT) A and B. Scores of the TMT B were corrected for age, sex, educational level, and TMT A-score (TMT part B given A t scores, TMT_B/A) using normative data reported from literature.³¹ Participants who were unable to finish the TMT part A or B in time (< 300 seconds) were scored as the lowest t score in the cohort, which was equal to 10. The following cardiovascular risk factors (CVRFs) were recorded: hypertension, hypercholesterolemia, diabetes type 1 or 2, and pack years of smoking, as defined for the DiViNAS cohort elsewhere.^{18, 28}

Brain MRI was performed on a single 3T MRI-system (Philips Diamond Select Achieva 3.0 [TX], Philips Medical Systems, Best, The Netherlands) using a 32-channel head coil, and included the following sequences: 3D T1-weighted images, T2-weighted images, fluid-attenuated inversion recovery, susceptibility-weighted images, and diffusion-weighted imaging with 30 directions. Detailed acquisition parameters and quantification methods are published elsewhere.^{18, 28} Small vessel disease (SVD) neuroimaging outcomes were peak width of the skeletonized mean diffusivity (PSMD), WMH volume (WMHv), brain volume, CMB count, lacune volume (LV), and a semiquantitative scale for PVS burden.³² WMHv, LV, and brain volumes were assessed following STRIVE-2 consensus criteria³³ and divided by the intracranial volume (ICV) to calculate normalized WMHv (nWMHv), normalized LV (nLV), and brain parenchymal fraction (BPF). PSMD was calculated from pre-processed data using a publicly available script (https://github.com/miac-research/psmd, version 1.8.3) with standard parameters.³⁴

Statistical Analysis

Statistical analysis was performed in R version 4.4.0. Variables with a normal distribution were reported using the mean ± standard deviation (SD); variables with a skewed distribution were reported using the median ± interquartile range (IQR). Multivariable linear mixed effects regression analyses were performed to investigate the association among PSMD (n = 192), nWMHv (n = 208), BPF (n = 209), nLV (n = 189), CMB count (n = 188), PVS score (n = 175), and TMT_B/A (n = 211) as dependent variables and the NOTCH3 score as independent variable using the lme4 package. The PVS score was treated as a continuous variable to facilitate comparisons with other continuous variables, but a sensitivity analysis was performed with (quartiles of) the PVS score as an ordinal outcome. The association between the NOTCH3 score and lifetime stroke probability (n = 212) was assessed with multivariable Cox regression analysis using the coxph package. In all multivariable analyses, age (excluded in Cox regression), sex, and CVRFs were included as covariates next to the NOTCH3 score, with batch numbers included either as a random effect or as a cofactor for the Cox regression models. For the outcomes that were significantly associated with the NOTCH3 score, an additional multivariable analysis was performed including NOTCH3 variant risk category.

To investigate modifiers of the NOTCH3 score, multivariable linear mixed effects regression analysis was performed with the following independent variables: NOTCH3 variant risk category, age, sex, CVRFs, with batch numbers as a random effect. All analyses were performed for the neuroimaging data collected at baseline (the main dataset) and were confirmed independently for the data collected at 2-year follow-up (the replication dataset), except for the PVS score, which was only available at baseline. To compare the NOTCH3 score in brain vessels of patients with HR- and MR-NOTCH3 variants, a Wilcoxon signed rank test was performed; for testing the association between age at the time of death and the NOTCH3 score, the Spearman's rank correlation was used. To investigate whether there was a significant 2-year change in the NOTCH3 score (n = 110; Supplementary Table S1), a linear mixed model (LMM) was created with follow-up time as a continuous fixed effect with crossed random intercepts for ID and batch number, and compared with a null model using a likelihood-ratio test.

Mediation analyses were performed between NOTCH3 variant risk category, the NOTCH3 score and clinical and neuroimaging outcomes. This was only performed for outcomes that were significantly associated with the NOTCH3 score in a multivariable model after stratification for NOTCH3 variant risk category. In each model, we included age, sex, and CVRFs as fixed effects, and batch numbers as a random effect. CVRFs, age, and sex were included as confounders as these have been consistently reported to be key modifiers of CADASIL disease severity.^{13, 18, 23, 35} A Quasi-Bayesian Confidence Intervals approach with 10,000 simulations was performed using the mediation R package. Sensitivity analyses on the mediation effects for violations of the assumption of no unmeasured confounding were conducted using the medsens function in the mediation package. Because this function does not support LMMs, mediation analyses were also performed using linear regression models; plots of the average mediation effects as a function of the sensitivity parameter rho³⁶ are shown in Supplementary Figure S3.

For the analyses of the NOTCH3 score of distinct NOTCH3 variants, the main and validation datasets were combined. To examine to what extent intervariant differences explained the additional variability in the NOTCH3 score, a likelihood-ratio test was performed between an LMM of the NOTCH3 score that only included NOTCH3 variant risk category, and an LMM with each EGFr domain as an additional cofactor in all patients with HR- and MR-NOTCH3 variants. The variant-specific NOTCH3 scores were only calculated for NOTCH3 variants present in 10 participants to ensure reliable estimates of averages, and were equal to the estimated marginal means (EMMs) per variant derived from an LMM with ID and batches as crossed random effects, and distinct NOTCH3 variant and sex as fixed effects, using the emmeans package. Each participant with a certain NOTCH3 variant was then assigned this variant-specific NOTCH3 score. For those outcomes that were significantly associated with the NOTCH3 score, LMMs and Cox regression analyses were performed to assess the association between the variant-specific NOTCH3 score and disease severity. Comparisons were performed between models to predict disease severity using the variant-specific NOTCH3 score and a participant's individual NOTCH3 score using Akaike Information Criteria (AIC). Differences in AICs of more than 2 were considered evidence that the lowest scoring model performed better.

To obtain plausible normal or homoscedastic residual distribution for linear mixed effects regression analyses, transformations were performed for PSMD and CMB count (natural logarithm), nWMHv (square root), and nLV (cube root). For all analyses, the NOTCH3 score was square root transformed and pack years were natural logarithm transformed. All continuous dependent and independent variables were standardized and scaled. Two-sided p values < 0.05 were considered statistically significant. Correction for multiple testing for neuroimaging and clinical outcomes and modifiers of the NOTCH3 score was performed using the Holm method, and reported as adjusted p value (p_adj).

Results

The Association Between the NOTCH3 Score and Neuroimaging and Clinical Outcomes

Descriptive characteristics of study participants are summarized in the Table 1. The NOTCH3 score (Fig 1B-G) was significantly associated with PSMD (β = 0.34, 95% confidence interval [CI] = 0.24 to 0.43, p = 6.0 × 10⁻¹¹, p_adj = 4.2 × 10⁻¹⁰), nWMHv (β = 0.38, 95% CI = 0.28 to 0.48, p = 5.4 × 10⁻¹², p_adj = 4.3 × 10⁻¹¹), nLV (β = 0.34, 95% CI = 0.22 to 0.46, p = 1.5 × 10⁻⁷, p_adj = 8.8 × 10⁻⁷), CMB count (β = 0.20, 95% CI = 0.08 to 0.32, p = 0.0017, p_adj = 0.0057), and PVS score (β = 0.22, 95% CI = 0.09 to 0.36, p = 0.0014, p_adj = 0.0057), but not with BPF (β = 0.017, 95% CI = −0.076 to 0.11, p = 0.72, p_adj = 1). The statistical inference of the NOTCH3 score remained significant when considering PVS score as an ordinal outcome (Supplemental Results; Data S1).

TABLE 1. Descriptive Characteristics

Number of Participants, n	212
NOTCH3 variant risk category, n (%)
HR-NOTCH3	117 (55.2)
MR-NOTCH3	88 (41.5)
LR-NOTCH3	7 (3.3)
Age, yr, mean (SD)	52.3 (12.2)
Male sex, n (%)	101 (47.6)
NOTCH3 score, median (range, IQR)	15.7 (0.86–55.9, 21.1)
Ischemic stroke, n (%)	71 (33.5)
Age of first stroke, yr, mean (SD)	50.9 (11.6)
Hypertension, n (%)	50 (23.6)
Hypercholesterolemia, n (%)	80 (37.8)
Diabetes type 1 or 2, n (%)	12 (5.7)
Pack years, median (IQR)	0.25 (10.1)

HR-NOTCH3 = high-risk NOTCH3 variant risk category; MR-NOTCH3 = moderate-risk NOTCH3 variant risk category; LR-NOTCH3 = low-risk NOTCH3 variant risk category; IQR = interquartile range.

The NOTCH3 score was significantly associated with a higher life-time risk of stroke (hazard ratio = 1.8, 95% CI = 1.4 to 2.4, p = 3.1 × 10⁻⁵, p_adj = 1.5 × 10⁻⁴; Fig 1H). There was no association between the NOTCH3 score and TMT_B/A (β = 0.024, 95% CI = −0.13 to 0.16, p = 0.74, p_adj = 1; Fig 1I). There was no association between age (range = 26–81 years) and the NOTCH3 score in skin (β = 0.021, 95% CI = −0.10 to 0.15, p = 0.75, p_adj = 1; Fig 1J) or brain vessels (p = 0.54; Fig 1K). There was no significant 2-year change in the NOTCH3 score (β = −0.078, 95% CI = −0.21 to 0.057, p = 0.25).

The Association between the NOTCH3 Score and NOTCH3* Variant Risk Categories*

Participants with an HR-NOTCH3 variant had a significantly higher NOTCH3 score than participants with an MR-NOTCH3 variant (β = 0.78, 95% CI = 0.54 to 1.0, p = 5.7 × 10⁻⁹, p_adj = 4.0 × 10⁻⁸; Fig 2A). In brain vessels, the NOTCH3 score in the HR-NOTCH3 group was significantly higher than in the MR-NOTCH3 group (p = 0.0091; Fig 2B). The association between age and the NOTCH3 score did not differ significantly between the HR- and MR-NOTCH3 groups (p = 0.14; Supplementary Fig S1). The NOTCH3 score was found to be a significant mediator of neuroimaging outcomes and lifetime stroke probability (Supplementary Fig S2; for a sensitivity analysis, see Supplementary Fig S3).

After stratification for NOTCH3 variant risk category, the association between the NOTCH3 score and neuroimaging outcomes and lifetime stroke probability remained significant, but not PVS score (Fig 2C–G). The associations between the NOTCH3 score and all neuroimaging markers were confirmed in the replication dataset, except for CMB count when stratifying for NOTCH3 variant risk category (see Supplemental Results; Data S1). Male participants had a lower NOTCH3 score than female participants in the main dataset (β = −0.39, 95% CI = −0.63 to −0.15, p = 0.0019, p_adj = 0.011; see Supplementary Fig S1), but this was not confirmed in the replication dataset. There was no association between CVRF and the NOTCH3 score (p_adj = 1 for all CVRF).

The NOTCH3 Score of Distinct NOTCH3* Variants and the Association With Disease Severity Outcomes*

The 39 distinct NOTCH3 variants of the DiViNAS participants were distributed over 19 EGFr domains (Fig 3 and Supplementary Table S1). Differences between EGFr domains accounted for a significant proportion of variability observed in the NOTCH3 score in the HR- and MR-NOTCH3 groups (p = 1.0 × 10⁻¹⁹). The average NOTCH3 score of several distinct NOTCH3 variants deviated from the average of their NOTCH3 variant risk category classification (see Fig 3). The most frequent NOTCH3 variants (present in n ≥ 10 participants) were the MR-NOTCH3 variant p.(Arg578Cys; n = 47), and the HR-NOTCH3 variants p.(Arg141Cys; n = 16), p.(Arg153Cys; n = 11), p.(Arg182Cys; n = 12), p.(Arg207Cys; n = 33), and p.(Cys1015Arg; n = 10). The variant-specific NOTCH3 scores of these 6 variants (Fig 3C, D) were significantly associated with PSMD (β = 0.50, 95% CI = 0.38 to 0.62, p = 6.3 × 10⁻¹³), nWMHv (β = 0.56, 95% CI = 0.45 to 0.67, p = 1.4 × 10⁻¹⁷), nLV (β = 0.59, 95% CI = 0.46 to 0.72, p = 4.2 × 10⁻¹⁴) (Fig 4A–C), CMB count (β = 0.30, 95% CI = 0.16 to 0.45, p = 6.7 × 10⁻⁵), PVS score (β = 0.30, 95% CI = 0.13 to 0.47, p = 5.6 × 10⁻⁴) and lifetime stroke probability (hazard ratio = 2.3, 95% CI = 1.6 to 3.2, p = 1.6 × 10⁻⁶). The variant-specific NOTCH3 score performed better than an individual participant's NOTCH3 score in predicting disease severity outcomes (Supplementary Table S4).

The variant-specific NOTCH3 score of p.(Arg207Cys) was observed to be notably lower than that of other HR-NOTCH3 variants. This was statistically significant (β = 1.1, 95% CI = 0.78 to 1.4 p = 4.9 × 10⁻¹¹), and the variant-specific NOTCH3 score of the p.(Arg207Cys) variant was comparable to that of MR-NOTCH3 variants (β = −0.079, 95% CI = −0.37 to 0.21, p = 0.60) (Fig 5A). Participants with the p.(Arg207Cys) variant also had a significantly lower PSMD, nWMHv, nLV, and lifetime stroke probability than participants with other HR-NOTCH3 variants (Fig 5B–E); CMB count did not significantly differ (see Supplementary Table S5 for coefficients and p values). The average NOTCH3 score of the p.(Arg578Cys) variant did not differ significantly from the average of the other MR-NOTCH3 variants, although there was a trend toward a lower NOTCH3 score in the p.(Arg578Cys) variant (reference group = p.[Arg578Cys], β = 0.33, 95% CI = −0.0001 to 0.66, p = 0.055).

Discussion

The findings of this study give novel insights into the interplay between cysteine-altering NOTCH3 variants, vascular NOTCH3^ECD protein aggregation, and NOTCH3-associated small vessel disease severity. We establish that (i) NOTCH3^ECD aggregation load in skin vessels is associated with neuroimaging outcomes and stroke risk, and (ii) that the average NOTCH3 score of distinct NOTCH3 variants is a predictor of disease severity, contributing to an improved genotype-based disease prediction.

NOTCH3^cys variants occur at an estimated population frequency of 1 in 300 and are associated with a highly variable small vessel disease severity, with CADASIL representing the severe end of the spectrum.^8-12 We previously showed that NOTCH3 variants can be classified into 3 risk categories for developing severe disease.²⁰ In this study, we demonstrate that the differences in disease severity between individuals with HR- and MR-NOTCH3 variants is mediated through differences in NOTCH3^ECD aggregation load. However, even within these risk categories and when accounting for cardiovascular risk burden, there remains a significant unexplained variability in disease severity. The results show that the NOTCH3 score can be used to more accurately predict disease severity associated with a particular NOTCH3 variant, paving the way for improved individualized genotype-based disease prediction in the clinic.

Our data suggest that NOTCH3^cys variants, of which over 370 have been reported,²⁰ are associated with differential NOTCH3^ECD aggregation load at the individual variant level, although there seems to be some topographical clustering along the NOTCH3 ectodomain of variants with similar NOTCH3 scores. This could be due to the 3-dimensional configuration of the NOTCH3 protein,²⁰ or the proximity to non-enzymatic NOTCH3 cleavage sites.^{37, 38} The amino acid that replaces, or is replaced by, a cysteine residue does not seem to be a major determinant of mutant protein properties.³⁹ Other factors may include expression levels of NOTCH3 and of extracellular matrix proteins known to be sequestered in NOTCH3^ECD aggregates.^{1, 40-42} Vascular NOTCH3^ECD aggregation is likely a complex interplay of these and other factors within the specific context of the vessel wall.

We did not find an association between CVRF burden and NOTCH3^ECD aggregation, and an inconsistent association with male sex. These known modifiers of CADASIL severity^{13, 18, 23, 35} therefore likely do not exert their effect via the NOTCH3^ECD aggregation pathway. There was no association of the NOTCH3 score with BPF or with TMT_B/A, in line with what was found previously for NOTCH3 variant risk category.¹⁸ In vitro assays in patient-derived inducible pluripotent stem cell 3D-vessel on chip models,⁴³ comparative multi-omics studies in patient tissues,⁴¹ as well as advances in artificial intelligence (AI)-driven protein 3D modeling tools, such as AlphaFold,⁴⁴ may contribute to our further understanding of NOTCH3^ECD aggregation and its modifiers.

Given the fact that CADASIL is a progressive disorder and it has been shown that the cerebrovascular NOTCH3 score increases with age in CADASIL mouse models,^{15, 26} we had expected to find an association between the NOTCH3 score and age, which was not the case. We cannot rule out that technical limitations or a bias in our cohort precluded the capture of such an association. On the other hand, these findings may imply that NOTCH3^ECD aggregation as measured with the NOTCH3 score reaches a plateau somewhere in the first 2 decades of life, as NOTCH3^ECD aggregation in skin vessels has been shown to be present in individuals as young as 19 years of age.^{16, 45-47} This hypothesis is further corroborated by the fact that we found no association between the NOTCH3 score in brain vessels and age at the time of death, whereas there was a significant difference in the NOTCH3 score of brain vessels between patients with HR- versus MR-NOTCH3 variants. Although we have previously shown an association between the NOTCH3 score in skin and brain,¹⁶ progression of NOTCH3^ECD aggregation in brain vessels likely differs from that in skin. The link between NOTCH3^ECD aggregation load and disease severity is probably manifold stronger if it could be measured in brain vessels using in vivo molecular imaging techniques. Development of such tools to measure cerebrovascular progression of NOTCH3^ECD aggregation may help to determine the optimal age to initiate treatment with disease modifying therapies targeting vascular NOTCH3^ECD aggregation.^{26, 27}

Both a strength and limitation of our study is that some NOTCH3 variants are prevalent in The Netherlands and are therefore over-represented in the DiViNAS cohort, such as the MR-NOTCH3 p.(Arg578Cys) and HR-NOTCH3 p.(Arg207Cys) variants. This provided enough statistical power to compare some distinct NOTCH3 variants with one another, but these over-represented variants prohibited the distinction between aggregation properties of distinct variants versus shared aggregation properties of variants within a specific EGFr domain. There were some variants which showed a striking deviation from the average NOTCH3 score of the NOTCH3 variant risk category in which they are (currently) classified, such as the p.(Arg1076Cys) variant in EGFr domain 27. The sample size of this variant was too low to draw any conclusions regarding potential re-classification of this variant, but the observation is intriguing, as EGFr 27 is located next to the HR-NOTCH3 EGFr domain 26. We could not study the NOTCH3 score in patients with non-cysteine altering^{5, 48} or low-risk NOTCH3 variants, as these were not represented or were under-represented in our study population. A sensitivity analysis of the mediation analyses showed that if there are unmeasured confounders, then these may strongly influence the estimated mediation effect of the NOTCH3 score. Although we cannot completely rule out such confounders, the results of the mediation analyses seem plausible, as we have included the main confounders previously reported in literature, and the mediation analyses showed consistent effects of the NOTCH3 score as a mediator for all associations between NOTCH3 variant risk category and the 5 disease outcomes. Finally, it is important to stress that, in its current form, the NOTCH3 score of a single patient cannot be used for individualized disease prediction nor is it suitable as a biomarker for clinical trials, as the current technique is not sensitive enough due to experimental and biological variation, including the number and type of vessels present in a skin biopsy section.

In conclusion, we show that the NOTCH3 score in skin vessels is directly associated with stroke probability and neuroimaging outcomes in patients with CADASIL, and that the NOTCH3 score may contribute to genotype-based risk classification for improved disease prediction.

Acknowledgments

The authors would like to thank all the patients with CADASIL and their families that have participated in our studies, and Ingrid Hegeman, Fabian-Kyle Lemmers, Anna Sczelkun, and Hannah Suikerbuik for their help in processing and sectioning of skin biopsy samples. This study was funded by Netherlands Organisation for Health Research and Development (91717325 and 09150182310003) and The Netherlands Brain Foundation (HA2016-02-03 and DR-2023-00420).

Author Contributions

M.N.C., G.G., J.W.R., and S.A.J.L.O. contributed to the conception and design of the study; M.N.C., G.G., R.J.H., K.L.D., M.R.G., B.G., M.N.W.W.-A., R.vD., M.D., J.W.R., and S.A.J.L.O. contributed to the acquisition and analysis of data; M.N.C., G.G., J.W.R., and S.A.J.L.O. contributed to drafting the text or preparing the figures.

Potential Conflicts of Interest

The authors declare no competing interests.

Open Research

Data Availability

The data that support the findings of this study are available upon reasonable request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Supporting Information

References

1Monet-Lepretre M, Haddad I, Baron-Menguy C, et al. Abnormal recruitment of extracellular matrix proteins by excess Notch3 ECD: a new pathomechanism in CADASIL. Brain 2013; 136: 1830–1845.
10.1093/brain/awt092
PubMed Web of Science® Google Scholar
2Joutel A, Vahedi K, Corpechot C, et al. Strong clustering and stereotyped nature of Notch3 mutations in CADASIL patients. Lancet 1997; 350: 1511–1515.
10.1016/S0140-6736(97)08083-5
CAS PubMed Web of Science® Google Scholar
3Joutel A, Andreux F, Gaulis S, et al. The ectodomain of the Notch3 receptor accumulates within the cerebrovasculature of CADASIL patients. J Clin Invest 2000; 105: 597–605.
10.1172/JCI8047
CAS PubMed Web of Science® Google Scholar
4Rutten JW, Haan J, Terwindt GM, et al. Interpretation of NOTCH3 mutations in the diagnosis of CADASIL. Expert Rev Mol Diagn 2014; 14: 593–603.
10.1586/14737159.2014.922880
CAS PubMed Web of Science® Google Scholar
5Muino E, Gallego-Fabrega C, Cullell N, et al. Systematic review of cysteine-sparing NOTCH3 missense mutations in patients with clinical suspicion of CADASIL. Int J Mol Sci 2017; 18:1964.
10.3390/ijms18091964
PubMed Web of Science® Google Scholar
6Zellner A, Scharrer E, Arzberger T, et al. CADASIL brain vessels show a HTRA1 loss-of-function profile. Acta Neuropathol 2018; 136: 111–125.
10.1007/s00401-018-1853-8
CAS PubMed Web of Science® Google Scholar
7Capone C, Cognat E, Ghezali L, et al. Reducing Timp3 or vitronectin ameliorates disease manifestations in CADASIL mice. Ann Neurol 2016; 79: 387–403.
10.1002/ana.24573
CAS PubMed Web of Science® Google Scholar
8Rutten JW, Hack RJ, Duering M, et al. Broad phenotype of cysteine-altering NOTCH3 variants in UK biobank: CADASIL to nonpenetrance. Neurology 2020; 95: e1835–e1843.
10.1212/WNL.0000000000010525
CAS PubMed Web of Science® Google Scholar
9Hack RJ, Rutten JW, Person TN, et al. Cysteine-altering NOTCH3 variants are a risk factor for stroke in the elderly population. Stroke 2020; 51: 3562–3569.
10.1161/STROKEAHA.120.030343
CAS PubMed Web of Science® Google Scholar
10Cho BPH, Nannoni S, Harshfield EL, et al. NOTCH3 variants are more common than expected in the general population and associated with stroke and vascular dementia: an analysis of 200 000 participants. J Neurol Neurosurg Psychiatry 2021; 92: 694–701.
10.1136/jnnp-2020-325838
PubMed Web of Science® Google Scholar
11Hack RJ, Gravesteijn G, Cerfontaine MN, et al. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy family members with a pathogenic NOTCH3 variant can have a normal brain magnetic resonance imaging and skin biopsy beyond age 50 years. Stroke 2022; 53: 1964–1974.
10.1161/STROKEAHA.121.036307
CAS PubMed Web of Science® Google Scholar
12Zhang R, Ouin E, Grosset L, et al. Elderly CADASIL patients with intact neurological status. J Stroke 2022; 24: 352–362.
10.5853/jos.2022.01578
PubMed Web of Science® Google Scholar
13Cho BPH, Harshfield EL, Al-Thani M, et al. Association of vascular risk factors and genetic factors with penetrance of variants causing monogenic stroke. JAMA Neurol 2022; 79: 1303–1311.
10.1001/jamaneurol.2022.3832
PubMed Web of Science® Google Scholar
14Rutten JW, Dauwerse HG, Gravesteijn G, et al. Archetypal NOTCH3 mutations frequent in public exome: implications for CADASIL. Ann Clin Transl Neurol 2016; 3: 844–853. https://doi.org/10.1002/acn3.344.
10.1002/acn3.344
CAS PubMed Web of Science® Google Scholar
15Rutten JW, Klever RR, Hegeman IM, et al. The NOTCH3 score: a pre-clinical CADASIL biomarker in a novel human genomic NOTCH3 transgenic mouse model with early progressive vascular NOTCH3 accumulation. Acta Neuropathol Commun 2015; 3: 89.
10.1186/s40478-015-0268-1
PubMed Web of Science® Google Scholar
16Gravesteijn G, Hack RJ, Mulder AA, et al. NOTCH3 variant position is associated with NOTCH3 aggregation load in CADASIL vasculature. Neuropathol Appl Neurobiol 2022; 48:e12751.
10.1111/nan.12751
CAS PubMed Web of Science® Google Scholar
17van den Boom R, Lesnik Oberstein SA, Ferrari MD, et al. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy: MR imaging findings at different ages—3rd-6th decades. Radiology 2003; 229: 683–690.
10.1148/radiol.2293021354
PubMed Web of Science® Google Scholar
18Hack RJ, Cerfontaine MN, Gravesteijn G, et al. Effect of NOTCH3 EGFr group, sex, and cardiovascular risk factors on CADASIL clinical and neuroimaging outcomes. Stroke 2022; 53: 3133–3144. https://doi.org/10.1161/STROKEAHA.122.039325.
10.1161/STROKEAHA.122.039325
CAS PubMed Web of Science® Google Scholar
19Rutten JW, Van Eijsden BJ, Duering M, et al. The effect of NOTCH3 pathogenic variant position on CADASIL disease severity: NOTCH3 EGFr 1-6 pathogenic variant are associated with a more severe phenotype and lower survival compared with EGFr 7-34 pathogenic variant. Genet Med 2019; 21: 676–682.
10.1038/s41436-018-0088-3
CAS PubMed Web of Science® Google Scholar
20Hack RJ, Gravesteijn G, Cerfontaine MN, et al. Three-tiered EGFr domain risk stratification for individualized NOTCH3-small vessel disease prediction. Brain 2023; 146: 2913–2927.
10.1093/brain/awac486
PubMed Web of Science® Google Scholar
21Cheng S, Xu Z, Bian S, et al. The STROMICS genome study: deep whole-genome sequencing and analysis of 10K Chinese patients with ischemic stroke reveal complex genetic and phenotypic interplay. Cell Discov 2023; 9: 75.
10.1038/s41421-023-00582-8
CAS PubMed Web of Science® Google Scholar
22Cho BPH, Jolly AA, Nannoni S, et al. Association of NOTCH3 variant position with stroke onset and other clinical features among patients with CADASIL. Neurology 2022; 99: e430–e439.
10.1212/WNL.0000000000200744
CAS PubMed Web of Science® Google Scholar
23Dupe C, Guey S, Biard L, et al. Phenotypic variability in 446 CADASIL patients: impact of NOTCH3 gene mutation location in addition to the effects of age, sex and vascular risk factors. J Cereb Blood Flow Metab 2023; 43: 153–166.
10.1177/0271678X221126280
CAS PubMed Web of Science® Google Scholar
24Min JY, Park SJ, Kang EJ, et al. Mutation spectrum and genotype-phenotype correlations in 157 Korean CADASIL patients: a multicenter study. Neurogenetics 2022; 23: 45–58.
10.1007/s10048-021-00674-1
CAS PubMed Web of Science® Google Scholar
25Mukai M, Mizuta I, Watanabe-Hosomi A, et al. Genotype-phenotype correlations and effect of mutation location in Japanese CADASIL patients. J Hum Genet 2020; 65: 637–646.
10.1038/s10038-020-0751-9
CAS PubMed Web of Science® Google Scholar
26Oliveira DV, Coupland KG, Shao W, et al. Active immunotherapy reduces NOTCH3 deposition in brain capillaries in a CADASIL mouse model. EMBO Mol Med 2023; 15:e16556.
10.15252/emmm.202216556
CAS PubMed Web of Science® Google Scholar
27Ghezali L, Capone C, Baron-Menguy C, et al. Notch3(ECD) immunotherapy improves cerebrovascular responses in CADASIL mice. Ann Neurol 2018; 84: 246–259.
10.1002/ana.25284
CAS PubMed Web of Science® Google Scholar
28Cerfontaine MN, Hack RJ, Gesierich B, et al. Association of NOTCH3 variant risk category with 2-year clinical and radiologic small vessel disease progression in patients with CADASIL. Neurology 2024; 102:e209310.
10.1212/WNL.0000000000209310
CAS PubMed Web of Science® Google Scholar
29Vandenbroucke JP, von Elm E, Altman DG, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): explanation and elaboration. PLoS Med 2007; 4:e297.
10.1371/journal.pmed.0040297
PubMed Web of Science® Google Scholar
30Bankhead P, Loughrey MB, Fernandez JA, et al. QuPath: open source software for digital pathology image analysis. Sci Rep 2017; 7: 16878.
10.1038/s41598-017-17204-5
PubMed Web of Science® Google Scholar
31Schmand B, Houx P, De Koning I. Normen van psychologische tests voor gebruik in de klinische neuropsychologie. Sectie Neuropsychologie Nederlands Instituut van Psychologen, 2012, published online: https://nip.nl/wp-content/uploads/2022/04/Handleiding-normen-Np-tests-2012.pdf.
Google Scholar
32Yao M, Herve D, Jouvent E, et al. Dilated perivascular spaces in small-vessel disease: a study in CADASIL. Cerebrovasc Dis 2014; 37: 155–163.
10.1159/000356982
PubMed Web of Science® Google Scholar
33Duering M, Biessels GJ, Brodtmann A, et al. Neuroimaging standards for research into small vessel disease-advances since 2013. Lancet Neurol 2023; 22: 602–618.
10.1016/S1474-4422(23)00131-X
PubMed Web of Science® Google Scholar
34Baykara E, Gesierich B, Adam R, et al. A novel imaging marker for small vessel disease based on skeletonization of white matter tracts and diffusion histograms. Ann Neurol 2016; 80: 581–592.
10.1002/ana.24758
PubMed Web of Science® Google Scholar
35Adib-Samii P, Brice G, Martin RJ, Markus HS. Clinical spectrum of CADASIL and the effect of cardiovascular risk factors on phenotype: study in 200 consecutively recruited individuals. Stroke 2010; 41: 630–634.
10.1161/STROKEAHA.109.568402
PubMed Web of Science® Google Scholar
36Imai K, Keele L, Yamamoto T. Identification, inference and sensitivity analysis for causal mediation effects. Stat Sci 2010; 25: 51–71.
10.1214/10-STS321
PubMed Web of Science® Google Scholar
37Young KZ, Lee SJ, Zhang X, et al. NOTCH3 is non-enzymatically fragmented in inherited cerebral small-vessel disease. J Biol Chem 2020; 295: 1960–1972.
10.1074/jbc.RA119.007724
CAS PubMed Web of Science® Google Scholar
38Zhang X, Lee SJ, Wang MM. Hydrolysis of a second asp-pro site at the N-terminus of NOTCH3 in inherited vascular dementia. Sci Rep 2021; 11: 17246.
10.1038/s41598-021-96679-9
CAS PubMed Web of Science® Google Scholar
39Lee SJ, Zhang X, Wu E, et al. Structural changes in NOTCH3 induced by CADASIL mutations: role of cysteine and non-cysteine alterations. J Biol Chem 2023; 299:104838.
10.1016/j.jbc.2023.104838
CAS PubMed Web of Science® Google Scholar
40Duering M, Karpinska A, Rosner S, et al. Co-aggregate formation of CADASIL-mutant NOTCH3: a single-particle analysis. Hum Mol Genet 2011; 20: 3256–3265.
10.1093/hmg/ddr237
CAS PubMed Web of Science® Google Scholar
41Zellner A, Müller SA, Lindner B, et al. Proteomic profiling in cerebral amyloid angiopathy reveals an overlap with CADASIL highlighting accumulation of HTRA1 and its substrates. Acta Neuropathol Commun 2022; 10: 6.
10.1186/s40478-021-01303-6
CAS PubMed Web of Science® Google Scholar
42Dupre N, Gueniot F, Domenga-Denier V, et al. Protein aggregates containing wild-type and mutant NOTCH3 are major drivers of arterial pathology in CADASIL. J Clin Invest 2024; 134:e175789.
10.1172/JCI175789
CAS PubMed Web of Science® Google Scholar
43Vila Cuenca M, Cochrane A, van den Hil FE, et al. Engineered 3D vessel-on-chip using hiPSC-derived endothelial- and vascular smooth muscle cells. Stem Cell Reports 2021; 16: 2159–2168.
10.1016/j.stemcr.2021.08.003
CAS PubMed Web of Science® Google Scholar
44Jumper J, Evans R, Pritzel A, et al. Highly accurate protein structure prediction with AlphaFold. Nature 2021; 596: 583–589.
10.1038/s41586-021-03819-2
CAS PubMed Web of Science® Google Scholar
45Lesnik Oberstein SA, van Duinen SG, van den Boom R, et al. Evaluation of diagnostic NOTCH3 immunostaining in CADASIL. Acta Neuropathol 2003; 106: 107–111.
10.1007/s00401-003-0701-6
CAS PubMed Web of Science® Google Scholar
46Tikka S, Mykkanen K, Ruchoux MM, et al. Congruence between NOTCH3 mutations and GOM in 131 CADASIL patients. Brain 2009; 132: 933–939.
10.1093/brain/awn364
PubMed Web of Science® Google Scholar
47Lesnik Oberstein SA, Maat-Schieman ML, Boon EM, et al. No vessel wall abnormalities in a human foetus with a NOTCH3 mutation. Acta Neuropathol 2008; 115: 369–370.
10.1007/s00401-007-0335-1
PubMed Web of Science® Google Scholar
48Ishiyama H, Kim H, Saito S, et al. Pro-hemorrhagic cerebral autosomal dominant Arteriopathy with subcortical infarcts and leukoencephalopathy associated with NOTCH3 p.R75P mutation with low vascular NOTCH3 aggregation property. Ann Neurol 2024; 95: 1040–1054.
10.1002/ana.26916
CAS PubMed Web of Science® Google Scholar

Volume98, Issue2

August 2025

Pages 273-285

Association Between Vascular NOTCH3 Aggregation and Disease Severity in a CADASIL Cohort – Implications for NOTCH3 Variant-Specific Disease Prediction